1060
J. Agric. Food Chem. 2009, 57, 1060–1064
Quantification of Phenyllactic Acid in Wheat Sourdough Using High Resolution Gas Chromatography-Mass Spectrometry LIAM ANTHONY MATTHEW RYAN,†,‡ FABIO DAL BELLO,† MICHAEL CZERNY,§ PETER KOEHLER,§ AND ELKE KARIN ARENDT*,† Department of Food Science, Food Technology and Nutrition, and National Food Biotechnology Centre, Cork, Ireland, and Deutsche Forschungsanstalt fu¨r Lebensmittelchemie, Garching, Germany
In this study, high-resolution gas chromatography-mass spectrometry (HRGC-MS) was successfully used to quantify the level of phenyllactic acid produced by Lactobacillus plantarum strains during sourdough fermentation. Investigation of samples collected during fermentation revealed that the production of phenyllactic acid occurs throughout the growth of L. plantarum in sourdough, but the highest production rate was observed during the logarithmic growth phase. The highest amount, that is, 33.47 mg of phenyllactic acid/kg of dough, was measured in sourdough fermented by the antifungal strain L. plantarum FST 1.7. Sourdoughs fermented by different L. plantarum strains contained different amounts of phenyllactic acid, thus indicating that the production is strain-dependent. Phenylacetic acid was also detected during sourdough analysis, thus showing that the HRGC-MS protocol developed is suitable for the detection not only of phenyllactic acid, but also of a broader range of phenolic acids that are highly relevant, but present in very low amounts in sourdough. KEYWORDS: Gas chromatography; mass spectrometry; phenyllactic acid; Lactobacillus; sourdough; stable isotope dilution assay
INTRODUCTION
Among the natural preservation systems, the use of lactic acid bacteria in the form of sourdough has a long tradition (1, 2). In the past, the preservative activity of sourdough lactic acid bacteria was mainly attributed to the production of lactic and acetic acid, with their concurrent pH drop, and only recently the range of antimicrobial metabolites produced by lactic acid bacteria has been studied in detail. These studies have led to the identification of a number of different compounds which inhibit the growth of spoilage organisms (3-6). Among the compounds produced by sourdough lactic acid bacteria, 2-hydroxy-3-phenyl-propionic acid, that is, phenyllactic acid, was recently shown to be a strong antifungal agent, with previous studies showing no activity against human and animal cell lines (7). Besides its antimicrobial properties, phenyllactic acid has been detected in a number of foods and it is one of the most abundant aromatic acids found in honey (8). When present at low concentrations, phenyllactic acid does not affect the odor of a food product, unlike other compounds like acetic acid (4). * Corresponding author. Tel.: +35321 4902064. Fax: +353 214270213. E-mail:
[email protected]. † Department of Food Science, Food Technology and Nutrition, University College Cork. ‡ National Food Biotechnology Centre, University College Cork. § Deutsche Forschungsanstalt fu¨r Lebensmittelchemie.
Therefore, phenyllactic acid represents a promising natural device for controlling contaminants in food systems such as bread. We have recently identified phenyllactic acid as one of the key compounds responsible for the antifungal activity of sourdough fermented by Lactobacillus plantarum strains, but the actual level produced was not measured (6, 9). By using HPLC, measured levels of about 30 mg of phenyllactic acid/kg were found in L. plantarum sourdoughs (10). More recently, liquid chromatography-mass spectrometry (LC-MS) was used to quantify phenyllactic acid levels in back slopped sourdoughs (11). The main drawback of these techniques is however their low resolution, and thus an approach to accurately measure the level of phenyllactic acid in food systems is still lacking. In this study, we have used high resolution gas chromatography-mass spectrometry (HRGC-MS) to accurately measure the level of phenyllactic acid produced by lactic acid bacteria during sourdough fermentation. Comparison of the levels produced by different L. plantarum strains revealed the straindependent production of phenyllactic acid in lactobacilli. The developed protocol allowed also the quantification of phenylacetic acid produced during sourdough fermentation. MATERIALS AND METHODS Chemicals. Phenyllactic acid was purchased from Bachem (Weil am Rhein, Germany). All other chemicals were purchased from Sigma (Munich, Germany).
10.1021/jf802578e CCC: $40.75 2009 American Chemical Society Published on Web 01/12/2009
Quantification of Phenyllactic Acid in Sourdough
Figure 1. Acid catalyzed conversion of phenyllactic acid and [13C2]-
phenylacetic acid (internal standard) to the respective ethyl esters in wheat sourdough prior to extraction and analysis by GC-MS. 9: [13C]-label. Lactobacillus Cultures. The antifungal strains L. plantarum FST 1.7 (LP 1.7) (6) and L. plantarum FST 1.9 (LP 1.9) (9), as well as L. plantarum DSM 20174T (LPTS) were routinely grown on mMRS5 medium (12) at 30 °C for 48 h under microaerophilic conditions. Sourdough Fermentation. Wheat sourdough fermentations were performed as described previously (6). Briefly, an overnight culture was used to inoculate (at 1% level) 80 mL of mMRS5 broth and was incubated overnight at 30 °C. Cells were harvested by centrifugation at 4000g for 10 min, washed twice, and resuspended in 40 mL of sterile tap water, to give a final concentration of ca. 5 × 109 CFU/mL of water. Six hundred grams of wheat flour, 599 mL of sterile tap water, and 1 mL of cellular suspension, giving a dough yield ((weight of flour + weight of water)/(weight of flour) × 100) of 200, were mixed to homogeneity for 1 min with a Kenwood mixer. Doughs were incubated at 30 °C for 48 h. Bacterial cell counts, pH, and total titratable acidity (TTA) were evaluated as previously described (6). Control fermentations with the addition of antibiotics as well as lactic and acetic acids were performed as previously described (9). These controls were performed to ensure that phenyllactic acid accumulation was due to metabolic activity of L. plantarum strains. Quantification of Phenyllactic Acid in Sourdough by HRGCMS. The level of phenyllactic acid present in sourdough was determined using an extraction from sourdough of phenyllactic acid and the stable isotopically labeled internal standard phenylacetic acid ([13C2]-phenylacetic acid). Initially 1 g of sourdough was spiked with the internal standard [13C2]-phenylacetic acid (1 mL; c ) 607 µg/mL in water). Acids were esterified as ethyl esters by mixing the sourdough with 10 mL of ethanol (99.9%) and 15 drops of sulfuric acid (97%) followed by incubation for 1 h at 80 °C under stirring at 200 rpm (Figure 1). After cooling, 20 mL of water was added. Each sample was extracted 5 times with 50 mL of dichloromethane. Sodium sulfate (50 g) was added and the mixture was stirred at 200 rpm for 10 min, filtered (0.45 µm), and the dichloromethane phase was then collected. After removal of the solvent at 50 °C, the residue was dissolved in 5 mL of dichloromethane and subjected to HRGC-MS analysis. HRGC-MS analysis was performed as previously described (13) using a gas chromatograph type HP 5890 series II (Agilent, Waldbronn, Germany) coupled to a MAT-95S mass spectrometer (Finnigan, Bremen, Germany) using a 30 m × 0.32 mm i.d., 0.25 µm, fused silica capillary DB-FFAP (J&W Scientific, Folsom, CA, USA). The flow rate of the carrier gas helium was 2 mL/min. Samples were applied at 40 °C by the cold-on-column technique (13). After 2 min, the temperature was increased at a rate of 6 °C/min to 230 °C and held at 230 °C for 5 min. Mass spectra were generated in the electron impact mode (MS/ EI) at 70 eV. The most intense mass traces for phenyllactic acid and [13C2]-phenylacetic acid were extracted from the mass chromatograms and used for quantitation. The response factor (mass to signal intensity ratio) of both compounds was calculated by analyzing phenyllactic acid/ [13C2]-phenylacetic acid mixtures in the following mass ratios: 0.013: 1, 0.02:1, 0.025:1, 0.033:1, 0.05:1, 0.066:1, 0.1:1, 0.2:1, and 1:1. To ensure that both phenyllactic acid and [13C2]-phenylacetic acid were extracted and ethyl esterified to the same extent in sourdough, known
J. Agric. Food Chem., Vol. 57, No. 3, 2009
1061
levels of each compound were added to dough and were analyzed as described above. The limit of accurate quantification of phenyllactic acid was defined as the lowest level at which phenyllactic acid could be accurately quantified in dough with good peak form and signal intensity (signal-to-noise ratio higher than 5:1). Accumulation of Phenyllactic Acid during Sourdough Fermentation. The HRGC-MS protocol described above was used to quantify the amount of phenyllactic acid accumulating during sourdough fermented with the antifungal strain LP 1.7. Sourdough samples were collected after 0, 8, 11, 24, and 48 h of fermentation. Since the production of specific antifungal compounds can vary among different strains of the same species (14), the production of phenyllactic acid was investigated in wheat sourdoughs fermented by two additional L. plantarum strains, that is, LP 1.9 and LPTS. Sourdough samples were collected after 0, 24, and 48 h of fermentation and investigated as described above. For all investigated doughs, three sourdough samples were taken from three independent fermentations at each time point, with means and standard deviations being determined for each set of data. RESULTS AND DISCUSSION
In the context of using lactic acid bacteria as a natural tool for improving the shelf life of food products, we have recently described the use of L. plantarum strains to produce sourdough which improves the shelf life of wheat bread (6, 9). Among the different compounds isolated, phenyllactic acid was indicated as a major compound responsible for the antifungal activity of L. plantarum sourdough. In this work we have successfully applied HRGC-MS to accurately measure the amount of phenyllactic acid produced by L. plantarum during sourdough fermentation. Sourdough Fermentation. Wheat sourdoughs were fermented with three different L. plantarum strains, and cell counts, pH, and TTA were determined. Overall, similar cell counts, pH, and TTA values were measured, independent from the strain used (data not shown). Lactobacillus cell counts of ca. 1 × 109 CFU/g of dough were reached after 11 h of fermentation, and the values were constant until the end of fermentation, that is, after 48 h. Analysis of the bacteria growing on mMRS5 agar containing bromocresol green (to allow visual differentiation among different strains) revealed that only one strain was responsible for the sourdough fermentation (data not shown). At the end of fermentation, the doughs were characterized by a pH of ca. 3.7 and TTA of ca. 14. The rate of acidification of the doughs (pH drop and TTA increase) did not differ significantly between the selected L. plantarum strains. Quantification of Phenyllactic Acid by Using HRGC-MS. The amount of phenyllactic acid produced by L. plantarum during sourdough fermentation was determined using HRGC-MS. In general GC-MS techniques are time-consuming and require convoluted sample preparation that involves metabolite extraction and derivatization to improve volatility. To overcome these pitfalls, we developed a simple and rapid sample preparation protocol, using a relatively short treatment time, involving the acid catalyzed ethyl esterification of phenyllactic acid. Phenylacetic acid was chosen as the internal standard because of its structural similarity to phenyllactic acid and because it is commercially available as a [13C2]-labeled compound. The use of a commercially available internal standard removed the need to synthesis an isotopically labeled internal standard. To render them volatile and make them soluble in organic solvents, phenyllactic acid and [13C2]-phenylacetic acid were esterified with ethanol in the sourdough by acid catalysis (Figure 1). Analysis of the EI spectra from phenyllactic acid and [13C2]phenylacetic acid showed signals at m/z 194 and m/z 166, which correspond to the molecular weights of the ethyl esters of
1062
J. Agric. Food Chem., Vol. 57, No. 3, 2009
Ryan et al.
Figure 2. Calibration curve of mixtures of phenyllactic acid and [13C2]-phenylacetic acid versus response factor.
Figure 3. Calibration curve of phenyllactic acid concentration versus peak area.
phenyllactic acid and [13C2]-phenylacetic acid, respectively. Phenyllactic acid and [13C2]-phenylacetic acid showed elution times of 1359.6 and 1002 s, respectively. Besides the signals corresponding to the molecular weight, the most intense ions generated by EI ionization were m/z 176 and m/z 164 for phenyllactic acid and [13C2]-phenylacetic acid, respectively. Therefore, these fragment ions were used for quantification. To ensure the accuracy of the method developed, the response factors (mass to area ratios) obtained with fixed phenyllactic acid/[13C2]-phenylacetic acid mixtures were determined (Figure 2). The corresponding R2 value was 0.984. Both compounds were found during multiple esterifications to be extracted and ethyl esterified in equal yields from sourdough. After extraction with dichloromethane, the ethyl esterified samples were found to be below the detection limit in the residual sourdough. The limit of accurate quantification was also determined and found to be 10 µg of phenyllactic acid/kg of sourdough, that is, ca. 60 nM. This value is in good agreement with the lower limit of detection of metabolites reported for GC-MS (15). At concentrations lower than this value, poor peak form and signalto-noise ratio reduced the accuracy of quantification. Because of the time dependent increase of phenyllactic acid during sourdough fermentation response factors of 1.151, 1.098, 1.090, and 1.074 were used for phenyllactic acid quantification in sourdough fermented by LP 1.7 for 8, 11, 24, and 48 h, respectively. To evaluate the accuracy of phenyllactic acid quantification the peak area was graphed against the level of
phenyllactic acid quantified over the 48 h of fermentation by LP 1.7 (Figure 3) and the corresponding R2 value was 0.994. The response factors for sourdoughs fermented for 24 and 48 h were 1.208 and 1.145 or 1.104 and 1.095 for LP 1.9 or LPTS, respectively. The phenyllactic acid content of sourdough has been recently determined using HPLC (10) and LC-MS (11), and the reported values are in good agreement with the results of our study. However, even though both HPLC and LC-MS can be used for quantification of phenyllactic acid, HRGC-MS was found to offer significant improvements, in particular with respect to sample preparation time and overall sensitivity. While the HPLC method described involved an overnight treatment prior to extraction (10), in our study the ethyl esterification performed prior to the extraction required only 1 h, thus giving a significant improvement with respect to preparation time. Concerning the use of LC-MS for quantification of phenyllactic acid, direct injection into an ion trap LC-MS of phenyllactic acid solutions of 50 mg/mL as well as 50 mg/L were found to give a peak with a low signal-to-noise ratio (
